雙耳聽音與辨位之研究

人的聽感有兩個重要的部分，一是對聲音的感受，二是覺知音源的方位。本研究目的重點為雙麥克風搭配簡化人頭模型模擬真人聽覺的感知，包含雙耳量測到的人頭聲學特性與音源方位之辨識。首先了解人聽覺的特性以及影響頭部關係轉移函數(Head Related Transfer Function, HRTF)響應的主要構造，並根據這些構造來建立人頭模型，其中外耳的尺寸係參考ITU-T Rec.P.57，並以圓管作為耳道，透過LMS Virtual.Lab軟體的Acoustic模組，以邊界元素法分析單耳聲場，確立合適的耳道長度。頭部的尺寸係參考ITU-T Rec.P.58所建議的尺寸來設計，本研究製作了兩個頭部模型，第一個頭型的曲面完全在規範建議的範圍中，第二個頭型為簡易的橢球幾何，其x、y、z三軸的尺寸符合規範中的頭長、頭寬、頭高，透過實驗比較兩種頭型之HRTF差異，由量測結果得知兩種頭型的聲場極為類似，顯示人頭形狀可進一步簡化為橢球幾何。此外，量測成人生理耳結合橢球幾何人頭模型的HRTF，實驗結果在5kHz以下的頻率響應狀況與簡化人頭模型之轉移函數幾乎一致。在無響室中，維持相同量測條件，試驗簡化人頭模型與專業的聲學仿真人偶—HMS Ⅲ.L Artificial Head，右耳收錄訊號的頻率響應趨勢與聲壓位準，並討論兩者量測資料的差異，接著利用NI LabVIEW DSP 模組中的帶止濾波器改善由於模型材質與耳道結構造成聲壓響應過大的缺失。最後比較HMS Ⅲ.L Artificial Head與簡化人頭模型，兩者響度與銳度量測結果。當音源於四個方位角(θ=0°、90°、180°、270°)，量測場點於右耳，結果顯示兩者響度差在2.58 sone以內，銳度差低於0.2 acum。結合簡化人頭模型與雙麥克風量測以及心理聲學參數演算，建立雙耳聲音品質量測系統，此外，由雙耳量測聲音訊號的時間差計算音源角度，發展簡化仿真人偶聽音辨位的功能。

關鍵字

頭部關係轉移函數；心理聲學參數；邊界元素法；雙耳量測技術；音源定位

並列摘要

The purpose of this study is to develop a binaural measurement system to simulate the human hearing characteristics with dual microphones and simplified head model. This work contains two parts, one is about the human hearing sense, the other is identification of sound source position. For understanding the characteristics of auditory sense and the main response of head related transfer function, a simplified head model with the external ear size based on ITU-T Rec.P.57 were established. Besides, the simplified model with circular tube as ear canal. The acoustic module of LMS Virtual Lab was used to analyze the transfer function of simplified monaural through boundary element method, then figured out the optimized size for ear canal length. The size for head is designed by referring to ITU-T Rec.P.58. In this research, we designed two head models, the shape for one of them is designed in accordance with the specification; while the other one is a simple oval geometry, that three axes following the guidelines of specification. Comparing with the frequency response of HMS Ⅲ.L Artificial Head, we found high pressure gain at resonant frequency, which is attributed to canal cavity and material of the head. To improve the defect of high amplitude response, we utilized bandstop filter of NI LabVIEW DSP module to smooth the response of simplified head model. The psychoacoustic parameters such as loudness and sharpness had been calculated according to Zwicker and Fastl’s method, the calculations are compared with the experimental data of the four directions(θ=0°、90°、180°、270°) from HMS Ⅲ.L Artificial Head, showing the difference of loudness and sharpness between the two measurement system are within 2.58 sone and 0.2 acum respectively. In addition, the interaural time difference (ITD) can be calculated by dual-delay line algorithm for a low frequency sound, the location in front position of artificial head may be recognized; using this concept, the technology of an identification of sound source position was developed.

並列關鍵字

HRTFs ； Psychoacoustics ； BEM ； binaural technique ； sound localization

參考文獻

[1] Blauert, J. (1982). Spatial Hearing, The Psychophysics of Human Sound Localization. The MIT Press, Cambrige, Massachusetts.

[2] Goldstein, E. B. (2006). Sensation and Perception (7th, Seven Edition)

[3] Stevens, S. S., and Newman, E. B. (1936). The Localization of Actual Sources of Sound, America Journal of Psychology, 48, pp 297-306.

[4] Gulick, W. L., et al. (1989). Hearing: Physiological Acoustics, Neural Coding, and Psychoacoustics, Oxford.

[5] Shaw, E. A. G. (1974) The external ear, In: Handbook of sensory Physiology, vol. V/1. V

國際替代計量

雙耳聽音與辨位之研究

全文下載

主題瀏覽